What to do when ChatGPT gets your brand wrong
AI assistants get brands wrong when the sources they read are wrong, outdated, inconsistent or missing. You cannot edit an AI answer directly, but you can fix what it reads: correct your own pages and profiles, publish the current facts on sources assistants trust, and report errors through each assistant's feedback tools.
Updated · Citable
Step 1: find where the error comes from
| Symptom | Likely cause |
|---|---|
| Old price or plan | An old pricing page, a cached directory listing, a review article from last year |
| Wrong category or audience | Inconsistent descriptions across profiles |
| Confused with another company | Similar name, and not enough distinct facts about you |
| Features you do not have | A competitor comparison, or the model generalizing from your category |
| "I don't have information about that" | Too few sources mention you |
| Right with search, wrong without | Model memory from before your changes |
Ask with web search on and read the cited pages. Then ask without search: if the error only appears there, it is in the model's training data.
Step 2: fix what you control
- Update or redirect old pages (old pricing, retired features, previous product names).
- Update every profile and listing with your current fact sheet. See entity SEO.
- Add a clear statement of the corrected fact on your site, in text: "Pricing (updated October 2026): …".
- If you were confused with another company, add distinguishing facts to your homepage and about page: location, founders, founding year, what you do not do.
Step 3: correct the third-party sources
Email authors of articles with wrong facts; most will update a price or a description. Update directory and review profiles. Where a large source is wrong and slow to change, add new, dated sources with the current facts: a press release announcing the change (new pricing, a rebrand, a new product line) puts the correct version on many news pages at once.
Step 4: report it
ChatGPT, Gemini, Claude and Perplexity all have thumbs-down or report buttons on answers. Use them with a short correction and a link to the source. It will not fix the answer immediately for everyone, but it is the channel each vendor provides. Do not expect a direct correction path for brand facts; fixing sources is what works.
Questions people ask
Can I ask OpenAI to change what ChatGPT says about my company?
There is no general process for brand corrections. For personal data, OpenAI and others have privacy request forms. For business facts, fix the sources and use answer feedback.
How long until a correction shows up?
Search-based answers can change within days of sources being updated and recrawled. Answers from model memory change with new model versions.